Mouse models of cancer are a critical link between basic laboratory discovery and translation to human clinical trials. For target validation, in vivo models are often the ultimate arbiters of validity (e.g., for publication), and in drug development no new chemical entity advances without evidence of efficacy utilizing in vivo models. Although the predictive value of mouse cancer models is subject to debate, in vivo models do recapitulate many aspects of cancer biology that are absent from in vitro cell line models under non-physiological cell culture conditions. For pragmatic reasons, however, most large-scale target identification efforts strictly utilize cell line-based screening methods despite the clear limitatons of in vitro models. We hypothesize that incorporating in vivo models into primary target identification screens will facilitate the identification of clinically relevant targets, and may identify targets that would not be evident using only in vitro cell line screening approaches. Given the explosion of interest in epigenetic therapies, in this proposal we will use a pooled shRNA approach to comprehensively identify epigenetic vulnerabilities in T-cell acute lymphoblastic leukemia (T- ALL) and mixed lineage leukemia (MLL). Using a library of 3,040 shRNAs targeting 449 known and putative epigenetic regulatory proteins, we will simultaneously assess for dependencies using in vitro cell line models and in vivo orthotopic leukemia models. In the first Aim, we will surmount one of the major impediments to in vivo screening by deriving highly leukemogenic sub-lines representing T-ALL and MLL. There is an unmet need for better therapies for these poor prognosis leukemia subclasses. In the second Aim, we will use pooled shRNA assays, performed simultaneously in vitro and in vivo, to identify epigenetic vulnerabilities in these two subclasses of leukemia. Finally, in the third Aim, we will fully validte critical vulnerabilities, including achieving in vivo proof-of-concept for therapeutic targeting. Together, these studies utilize a convergence of innovative and emerging technologies to identify epigenetic vulnerabilities in leukemia. Success in these studies would establish a new paradigm for target discovery leveraging in vivo models at the inception of discovery campaigns. In addition to rapid elimination of false positive hits, this approach may identify new target classes that are opaque to standard in vitro screening methods.